Hhqqqqyinmathisimple.hashnode.dev·Apr 13 · 10 min readWhy Naive Bayes Still Outperforms Fancy Models When Data Is MessyOur fraud detection neural network had 12 layers, 2.3M parameters, and 68% precision. I replaced it with Naive Bayes — 0 layers, 847 parameters, 79% precision. Training time dropped from 4 hours to 1100
Hhqqqqyinmathisimple.hashnode.dev·Apr 12 · 8 min readWhy Your PCA Pipeline Works in Notebooks But Fails in ProductionWhy Your PCA Pipeline Works in Notebooks But Fails in Production Our image search model's precision dropped from 82% to 61% three days after deployment. The culprit? A PCA pipeline that worked perfect00
Hhqqqqyinmathisimple.hashnode.dev·Apr 10 · 7 min readThe Entropy vs Gini Debate No One Tells Engineers AboutI've reviewed 200+ pull requests where engineers spent hours debating criterion='gini' vs criterion='entropy' in scikit-learn's DecisionTreeClassifier. The accuracy difference? Usually less than 0.3%.00
Hhqqqqyinmathisimple.hashnode.dev·Apr 8 · 8 min readWhy Feature Scaling Matters: Three Cases Where the Same Data Gives Opposite ResultsThe same dataset. The same algorithm. Opposite predictions. Not because of a bug — because one version scaled the features and the other didn't. Feature scaling is one of those topics that gets mentio10
Hhqqqqyinmathisimple.hashnode.dev·Apr 7 · 9 min readMachine Learning Data Preprocessing: The Mistakes That Break Models Before TrainingThe model isn't the problem. Nine times out of ten, when a machine learning project falls apart — bad predictions, overfitting that training metrics didn't catch, inexplicable behavior on new data — t10
